On Tue, 2008-04-08 at 23:36 +0200, Johannes Berg wrote: > > > This version fixes the clobbering of r12 by the call to > > > trace_hardirqs_off() that was pointed out by BenH. > > > > > > Johannes, I'd appreciate your trying this version if/when > > > you get the chance. > > > > Thanks Dale, this version seems to work. I'll stress it a bit more > > later, but so far it has survived *much* longer than both previous > > versions. > > Hmm. Bad news. I got a crash in console_callback() again where > apparently r28 had a bogus value. That, however, doesn't make sense, so > maybe that register value was calculated from another register.
Possibly. How hard was it to reproduce ? Could it be a bug unrelated to lockdep ? Ben. _______________________________________________ Linuxppc-dev mailing list Linuxppc-dev@ozlabs.org https://ozlabs.org/mailman/listinfo/linuxppc-dev
